Open AccessDistributed corporate information systems optimization based on genetic algorithm and object model
Author(s) -
Svetlana Zemlyanskaya,
O. V. Chengar,
E. O. Savkova,
O. A. Shumaieva
Publication year - 2020
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1582/1/012019
Subject(s) - crossover , genetic algorithm , computer science , fitness function , object (grammar) , replication (statistics) , minification , algorithm , mutation , function (biology) , distributed object , mathematical optimization , optimization problem , artificial intelligence , mathematics , machine learning , evolutionary biology , biology , gene , programming language , biochemistry , statistics , chemistry
In paper object model of distributed corporate information system is developed in the form of interacting models of its typical components, which takes into account replication and distribution of data between nodes of computer network, characteristics of hardware and makes it possible to simulating DCIS of any structure. A new modification of the genetic algorithm for the DCIS optimization problem has been developed, in which combined multichromosomes have been used to represent the placement of database tables, applications and parameters of servers, network devices and communication channels, genetic operators of recombination, crossover and mutation have been developed. To optimize the DCIS functioning, together with a modified genetic algorithm, the object model is used, that calculates the fitness-function, forming the optimal configuration of DCIS data distribution and hardware parameters in real time for the performance criterion formulated as the minimization of the system response time. A serial scheme of implementation of the genetic algorithm was developed, which includes full or partial optimization of DCIS in accordance with the requirements. The approach proposed allows obtaining suboptimal solution to the DCIS optimization problem with deviation from the global optimum of no more than 5%.